hsrp groups using the same group number.

[francisco_1]

I understand on older IOS codes If the same hsrp group number is assigned to multiple standby groups, it creates a non-unique MAC address. Is this true on newer codes like 12.2(52)SE for 3750 & 3560?

Francisco.

[lgijssel]

HSRP uses the hsrp-group number in the virtual address like explained in the link below:

http://www.cisco.com/en/US/tech/tk648/tk362/technologies_tech_note09186a0080094afd.shtml#topic3

As far as I know, this has not changed and I also cannot see any need for it because the hsrp-group number must always be unique on a subnet.

As long as this is true, all mac addresses will also be unique on their respective subnets.

It is a common misunderstanding that mac addresses (which is what an hsrp address actually is) need to be unique throughout a network.

That's simply not true, they only need to be unique within a vlan. Hence it is perfectly safe to use the same hrsp group number on different vlans.

As long as there is no layer2 interconnection, everything will keep working perfectly.

Should you experience address collision between hrsp-groups, then there must be a layer2 backdoor connection between subnets.

This is then a problem which must be fixed, not a shortcoming of hsrp.

regards,

Leo

[Giuseppe Larosa]

Hello Francisco,

multilayer switches allow for HSRP group reuse on different Vlans/SVIs

This is seen not as a problem but as a scalability workaround in order to overcome some limitations of low end multilayer switches that didn't support more then 16 groups in some cases.

About the MAC address reuse: this is not a problem in multilayer switches as they are able with TCAM tables to track the Vlan where the frame has been received so that the packet is processed by the correct SVI.

Said this, a potential for problem is still present in case two different broadcast domains are joined by accident.

In this case the Vlan multiplexing is not enough to distinguish in context frames and out of context frames.

This can cause high cpu usage and I have seen this also on C6500 with Sup 720 running 12.2(18)SXF code.

A possible workaround for this is the following:

reuse the same HSRP group number on different Vlans but use a different MD5 authentication on each, so that out of context frames ( in the case described above) are early discarded.

Hope to help

Giuseppe

[Mohamed Sobair]

Hi,

The HSRP Group number defines the last byte of the virtual mac-address of the HSRP. This mac-address should be unique withing a broadcast domain. and is used for the local communication between HSRP members. should be there any conflict within this address or different Mac address between HSRP members (virtual Mac), then there will be no HSRP for that Group.

HTH

Mohamed

[MichaelTanner]

Hello,

How would multiple SVI's in a single HSRP Group effect failover? You would think an event within a VLAN could cause everything to failover in the HSRP Group including spanning tree loop, or even preempt on an SVI.